A Designer’s Guide to using AI to turbocharge creativity

Have you tried using AI to augment your workflow? If not, you’re in luck — I’m going to outline my end-to-end process for using AI to augment my workflow.

• Use ChatGPT to refine or expand upon your ideas, brainstorm anything
• Use Playground (or Midjourney, or DALLE-3) to generate images and storyboards.
• Use Runway (or Kaiber, Pika Labs) to generate videos and animations.
• Put it all together in Photoshop, Figma, After effects, or your editing software of choice.

Here’s the entire process (don’t get overwhelmed, I’m gonna walk through it step by step!):

Idea generation

First things first — you need ideas. How to create ideas is a relatively simple, age-old process. With GPT-4, it’s become 1,000x faster and easier.

Try this:

Let’s brainstorm ideas for __________ (what do you want to do?)
be sure to ______ (extra context or guidance)

Remember that it’s helpful to use GPT as a “thought partner”, rather than an “answer giver”. Use it to address the blank slate problem- and remix your own ideas into what you get back. Try it out now, GPT 3.5 is free to use.

With the release of Dalle-3, ChatGPT can generate ideas AND images in the same conversation seamlessly.

Image generation with Playground

There are many services that you can use to generate images using Stable Diffusion. The most popular one is Midjourney, however, my personal favorite is Playground (I am a paying customer of Playground).

Playground.ai has a great interface, and is my preferred stable diffusion tool.

I prefer Playground because it has the best UX of any stable diffusion product on the market right now. Specifically:

1. The product’s interface and core interaction loop make it SUPER EASY to explore and ITERATE through your ideas, until you get what you’re looking for, and then you can refine it from there.
2. The ‘filters’ feature makes it really simple to focus on the content of what you’re trying to create, rather than fiddling with trying to write the perfect prompt.
3. The built-in social features make it easy to browse images and find inspiration, and even remix others creations directly with one click.

Remember when prompting these systems, KEEP IT SIMPLE (less is more) and remember that it takes a bit of time to get a feel for what works to get the results you want.

Once you have some sweet images… it’s time to get animated.

Video generation with Runway

While the idea of text to video is literally magic, in my experience, we can get much better quality results starting from an image (that you already love) and animating that image. Starting from scratch gets you weird AI-looking things. But if you start with a gorgeous image, you’ll get a gorgeous video.

Runway is the best in the biz right now. They just released a huge update and get the best results of any off-the-shelf tool on the market right now.

Here’s why I love Runway:

1. Runway can perform Image to Video, Video to Video, or Text to Video.
2. They let you “extend” videos to go up to 16 seconds!
3. Their interface is easy to understand and forgiving, lets you go back/edit seamlessly so it’s easier to experiment without having to fully start over (ITERATION IS KEY!)

I have a second favorite: Kaiber. Their UX is kinda glitchy (cough cough, would love to chat) BUT they make it easy to get cool results on account of their filters (similar to Playground).

Pro tip: Use Stock Footage

Another technique is to start with stock footage. It sounds silly, I know, but royalty-free free stock downloads are a thing. You can do some amazing things with video to video transformation available in both runway and kaiber).

Footage from www.pexels.com

Once you have your video pieces, it’s time to stitch them together…

Edit Composite

At this point, you have these sweet video clips and AI-generated images. Use the editing software of your choice (Figma (for still images), Photoshop, After Effects, or Runway (runway has a video editing component, too!)) to combine your pieces into a finished masterpiece.

My process using Photoshop. Video goes at the bottom, with other static layers and textures on top.

I prefer to use ol’ reliable (Photoshop) for this job because:

1. Photoshop is simply the most robust image editor out there
2. …which also lets you combine hyper-detailed raster editing with video/gif animation frames
3. …and lets you control the gif output quality in nuanced ways to manage file size.

Yes, Photoshop can be slow at times (shakes fist at Adobe) but I haven’t found a better replacement. If you find a better way to combine .pngs and .mp4’s into .gif files please let me know.

You can loop a video and put some layers on top to create something truly breathtaking.

This is how I produced the album artwork for the High Frequency collections

And that’s it!

Export as GIF — and voila! You’ve created a one-of-a-kind masterpiece that blends the power of AI with your own creativity.

Here are some examples of things I’ve created this year using this exact workflow. I hope you enjoy them as much as I do!

Thanks for reading! If you found this useful, pump the algorithm by giving this post a clap, share, like, whatever — and if you make something—share it far and wide!! It’s still early days with this technology and we’re all figuring it out together.

Questions, comments? Shoot me a note on Twitter/X and I’ll get back to you! I love this stuff and am happy I finally am getting around to writing about it. ✌️❤️🤖